Characterization of a novel organelle in Toxoplasma gondii with similar composition and function to the plant vacuole

Authors

  • Kildare Miranda,

    1. Center for Tropical and Emerging Global Diseases and Department of Cellular Biology, University of Georgia, Athens, GA 30602, USA.
    2. Instituto de Biofisica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Rio de Janeiro 21941, Brazil.
    Search for more papers by this author
    • Both authors contributed equally.

  • Douglas A. Pace,

    1. Center for Tropical and Emerging Global Diseases and Department of Cellular Biology, University of Georgia, Athens, GA 30602, USA.
    Search for more papers by this author
    • Both authors contributed equally.

  • Roxana Cintron,

    1. Center for Tropical and Emerging Global Diseases and Department of Cellular Biology, University of Georgia, Athens, GA 30602, USA.
    Search for more papers by this author
  • Juliany C. F. Rodrigues,

    1. Center for Tropical and Emerging Global Diseases and Department of Cellular Biology, University of Georgia, Athens, GA 30602, USA.
    2. Instituto de Biofisica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Rio de Janeiro 21941, Brazil.
    Search for more papers by this author
  • Jianmin Fang,

    1. Center for Tropical and Emerging Global Diseases and Department of Cellular Biology, University of Georgia, Athens, GA 30602, USA.
    Search for more papers by this author
  • Alyssa Smith,

    1. Center for Tropical and Emerging Global Diseases and Department of Cellular Biology, University of Georgia, Athens, GA 30602, USA.
    Search for more papers by this author
  • Peter Rohloff,

    1. Center for Tropical and Emerging Global Diseases and Department of Cellular Biology, University of Georgia, Athens, GA 30602, USA.
    Search for more papers by this author
  • Elvis Coelho,

    1. Instituto de Biofisica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Rio de Janeiro 21941, Brazil.
    Search for more papers by this author
  • Felix De Haas,

    1. FEI Nanoport, Eindhoven, The Netherlands.
    Search for more papers by this author
  • Wanderley De Souza,

    1. Instituto de Biofisica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Rio de Janeiro 21941, Brazil.
    Search for more papers by this author
  • Isabelle Coppens,

    1. Department of Molecular Microbiology and Immunology, Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD 21205, USA.
    Search for more papers by this author
  • L. David Sibley,

    1. Department of Molecular Microbiology, Washington University School of Medicine, St. Louis, MO 63110, USA.
    Search for more papers by this author
  • Silvia N. J. Moreno

    Corresponding author
    1. Center for Tropical and Emerging Global Diseases and Department of Cellular Biology, University of Georgia, Athens, GA 30602, USA.
    Search for more papers by this author

E-mail smoreno@cb.uga.edu; Tel. (+706) 542 4736; Fax (+706) 542 9493.

Summary

Toxoplasma gondii belongs to the phylum Apicomplexa and is an important cause of congenital disease and infection in immunocompromised patients. Like most apicomplexans, T. gondii possesses several plant-like features, such as the chloroplast-like organelle, the apicoplast. We describe and characterize a novel organelle in T. gondii tachyzoites, which is visible by light microscopy and possesses a broad similarity to the plant vacuole. Electron tomography shows the interaction of this vacuole with other organelles. The presence of a plant-like vacuolar proton pyrophosphatase (TgVP1), a vacuolar proton ATPase, a cathepsin L-like protease (TgCPL), an aquaporin (TgAQP1), as well as Ca2+/H+ and Na+/H+ exchange activities, supports similarity to the plant vacuole. Biochemical characterization of TgVP1 in enriched fractions shows a functional similarity to the respective plant enzyme. The organelle is a Ca2+ store and appears to have protective effects against salt stress potentially linked to its sodium transport activity. In intracellular parasites, the organelle fragments, with some markers colocalizing with the late endosomal marker, Rab7, suggesting its involvement with the endocytic pathway. Studies on the characterization of this novel organelle will be relevant to the identification of novel targets for chemotherapy against T. gondii and other apicomplexan parasites as well.

Introduction

Toxoplasma gondii is a protist parasite that causes widespread infection in humans and has been recognized as a major opportunistic pathogen of immunocompromised patients. Additionally, first time infection with T. gondii of pregnant women poses a significant risk to the developing fetus. As a member of the phylum Apicomplexa, T. gondii possesses a distinct apical complex consisting of different types of secretory organelles, such as micronemes, dense granules and rhoptries, these latter organelles being acidic (Shaw et al., 1998; Dubremetz, 2007). T. gondii also contains acidocalcisomes, which are rich in calcium, pyrophosphate and polyphosphate and are acidified by a membrane-bound vacuolar proton pyrophosphatase (V-H+-PPase, or TgVP1) (Luo et al., 2001; Drozdowicz et al., 2003). In T. gondii, V-H+-PPase was found also in a large vacuolar compartment of tachyzoites that undergoes dynamic changes during invasion of host cells (Drozdowicz et al., 2003). An association of the V-H+-PPase with the endosomal/lysosomal pathway was proposed because the enzyme was found to label an endosome-associated compartment (‘VP1 compartment’) of the secretory pathway. This VP1-positive compartment accumulates microneme protein 2 (MIC2) when the propeptide of the non-anchored microneme protein 2-associated protein (M2AP) is deleted (Harper et al., 2006).

Large vacuolar compartments have been frequently observed in T. gondii. Interestingly, the protease cathepsin B localizes to a ‘large endosomal vacuole’ in tachyzoites, distinct from the rhoptries, the primary location of this protease (Que et al., 2002). TgRab51, a homologue of human Rab5, was observed to partially colocalize with TgVP1 (Harper et al., 2006) in similarly described ‘large lucent vacuoles’ that appear ‘empty’ (Robibaro et al., 2002). In addition, FITC-labelled heparin was observed in large ‘round vesicles’ spanning several optical sections of the tachyzoites ‘sometimes collectively occupying more than 50% of the parasite’ (Gross et al., 1993; Botero-Kleiven et al., 2001). These ‘large vacuoles’ have been referred to as pre-rhoptries or forming rhoptries (Robibaro et al., 2002) but a true link with the rhoptry compartment has never been established. In fact, the absence of specific markers has precluded the identification of these large vacuoles as lysosomes, late endosomes or lysosome-related organelles.

Although disputed (Hunter et al., 2007; Olbrich et al., 2007), there is experimental evidence indicating the presence of two functionally distinct vacuoles in plant cells: a lytic vacuole and a protein-storage vacuole (Paris et al., 1996; Jiang et al., 2002). These vacuoles are served by separate vesicular pathways from the Golgi: clathrin-coated vesicles for the lytic vacuole and dense vesicles for the protein-storage vacuole (Paris et al., 1996). Plant lytic vacuoles contain hydrolytic enzymes and function as digestive organelles similar to lysosomes in animal cells (Boller and Kende, 1979). All plant vacuoles contain the vacuolar H+-ATPase, the V-H+-PPase and tonoplast intrinsic proteins (TIP)-like aquaporins (Martinoia et al., 2007).

We describe a novel organelle in T. gondii, with broad similarities to the plant vacuole that contains transporters usually found in the plant vacuole membrane (tonoplast). In addition, this novel organelle appears to have homeostatic and calcium storage functions in the extracellular stage of the parasite. The relevance of this organelle in T. gondii extracellular parasites is discussed in relation to the known functions of the plant vacuole.

Results

A large vacuole in extracellular tachyzoites labels with antibodies against a vacuolar-H+-pyrophosphatase (TgVP1), a cathepsin L (TgCPL) and an aquaporin (TgAQP1)

All plant vacuoles contain the V-H+-pyrophosphatase (Rea and Poole, 1986), an enzyme that was originally described in Rhodospirillum rubrum (Baltscheffsky and von Stedingk, 1966; Moyle et al., 1972), and more recently in acidocalcisomes of several microorganisms (Scott et al., 1998; Docampo et al., 2005), including T. gondii (Rodrigues et al., 2000). In order to characterize and study the localization of this enzyme in T. gondii we produced antibodies against two distinct peptides in the sequence of the T. gondii V-H+-PPase (TgVP1). Immunofluorescence analysis (IFA) of extracellular tachyzoites with one of these antibodies shows labelling of a large vacuolar structure also clearly observable by differential interference contrast (DIC) microscopy (Fig. 1A and B). This enzyme was previously localized to the acidocalcisomes (Rodrigues et al., 2000), and vesicles labelled with the antibody are observed in all preparations with both antibodies (see Figs 1B and S1A, arrows). Partial colocalization of both antibodies is shown in Fig. S1A. The specificity of these antibodies was further confirmed by comparing their localization in cells expressing an epitope-tagged version of TgVP1 (Fig. S1B). Western blot analysis of tachyzoite total lysates from RH (wild-type strain) and TgVP1-tagged overexpressing parasites also confirmed antibody specificity (Fig. S1C). A band of 80 kDa was detected in cell overexpressing the myc-tagged protein using the anti-myc antibody (Fig. S1C, lane 1) and a second band of 72 kDa, corresponding to the endogenous protein, was detected in total lysates using the anti-TgVP1 antibody (Fig. S1C, lane 3). A size discrepancy between the expected (∼89 kDa) and the observed molecular mass has been reported for the V-H+-PPases of plants (Sarafian et al., 1992; Kim et al., 1994) and trypanosomatids (Hill et al., 2000; Lemercier et al., 2002). It could be attributed either to the anomalous migration of hydrophobic proteins on SDS gels (Maddy, 1976) or to partial degradation. We also show that this large vacuolar compartment labels with anti-TgVP1 by immunoelectron microscopy (IEM) (Fig. 1C, right panel). Figure 1C, left panel, shows labelling of a smaller vacuole of the size of an acidocalcisome (about 200 nm) (Luo et al., 2001).

Figure 1.

Antibody specific to the vacuolar proton pyrophosphatase (TgVP1) consistently labels a vacuolar structure in extracellular tachyzoites.
A. Extracellular tachyzoites were incubated in Ringer's buffer, fixed and stained with anti-TgVP1 antibody (1:4000) and DAPI (blue). Left panel shows the DIC images of three parasites with arrows pointing at a large vacuolar compartment clearly visible. The right panel shows strong labelling with anti-TgVP1 in a large vacuole in the three parasites.
B. Reaction with anti-TgVP1 in tachyzoites overexpressing TgVP1 (TgVP1-OE). Arrow points to an acidocalcisome.
C. Immunogold electron microscopy labelling with anti-TgVP1 antibody of a small empty vacuole with the size of an acidocalcisome (left) and a large vacuolar compartment (right). The antibody used was the purified rabbit anti-serum generated against the peptide: SGKNEYGMSEDDPRN. Bars: A and B = 5 µm, C = 200 nm.

A large vacuolar compartment was also observed in sectioned tachyzoites by transmission electron microscopy (Fig. 2A–C), or by electron tomography (Fig. 2D). This large compartment contained single membrane-bounded vesicles of diverse size and appearance and was occupied by a less electron dense material than that present in the cytosol (Fig. 2A–C). Some of these vesicles were more electron-lucent and contained very electron-dense material (Fig. 2B, inset), as is typical of acidocalcisomes (Docampo et al., 2005). In some cases it was possible to observe typical acidocalcisomes (empty vacuoles containing electron-dense material) in physical contact with this compartment appearing to fuse and even internalize (Fig. 2B–D, arrow in D).

Figure 2.

Transmission electron microscopy of recently released tachyzoites.
A.–C. Thin sections of whole tachyzoites showing a large vacuole. Arrow in (B) points to an acidocalcisome-like organelle. The inset shows the enlargement of the vacuole containing this structure. The presence of internal vesicles is also evident inside the large vacuole in (C). R, rhoptry; DG, dense granule; N, nucleus; A, apicoplast. Scale bars: A, B = 1 µm; C = 0.5 µm.
D.–G. 3-D reconstruction of a T. gondii tachyzoite by electron tomography. The image in (D) shows a profile of a tachyzoite within the reconstructed volume showing intracellular structures, an acidocalcisome and a large vacuole containing internal vesicles. The 3-D models (E–G) show a segmented plasma membrane (golden yellow), the mitochondrion (light blue), acidocalcisomes (orange) and a large vacuole (green). Different tilted views of the models are presented where it is possible to observe internal vesicles (dark gold). Arrow in (D) shows point of contact of an acidocalcisome with the large vacuole.

Plant lytic vacuoles are characterized by a high content of hydrolytic enzymes, such as proteases, and it has been proposed that amino acid recycling by protein degradation is a major function of the plant vacuole (Muntz, 2007). A cathepsin L-like enzyme (TgCPL) with homology to the previously described vacuolar aleurain from barley (Rogers et al., 1985) has been shown to localize to ultrastructural empty vesicles in T. gondii (Huang et al., 2009; see also Larson et al., 2009). We investigated the potential colocalization of TgCPL with TgVP1. As shown in Fig. 3A, TgVP1 and TgCPL show colocalization to the same organelle with TgCPL occupying the interior of a large vacuole while antibodies against TgVP1 labelled its membrane (Movie S1 shows that both markers are in the same compartment). In other cells TgCPL gave a strong reaction in the periphery of the vacuole (see for example Fig. 3B, TgCPL and merge). This is possible because the propeptide of TgCPL has a transmembrane domain and could be inserted in the membrane. It is possible that the two localizations could represent different stages of maturation of the enzyme, at first membrane-associated and then cleaved to be in the lumen. The accompanying article by Parussini et al. (2010) shows in detail the biochemical characterization and subcellular distribution of TgCPL.

Figure 3.

Colocalization of TgVP1 and TgCPL as observed by IFA and immunogold electron microscopy. Tachyzoites were fixed and stained with rabbit anti-TgVP1 antibody (1:4000) (green), and mouse anti-CPL antibody (1:400) (red) and DAPI (blue). Both antibodies localize to the same compartment (merge and overlay). Anti-TgVP1 labels the membrane of a large vacuole [green and merge in (A) and (B)]. Anti-TgCPL is also observed at the membrane of the large vacuole [red and merge in (B)]. Scale bars: A = 10 µm; B = 3 µm. (C) Immunogold electron microscopy showing colocalization of TgVP1 (rabbit serum) with TgCPL. Colocalization of TgVP1 (5 nm gold particles, arrows) with TgCPL (10 nm gold particles) to the membrane of a large vacuole (C and D) or to internal vesicles (C). Scale bars = 0.1 µm.

The colocalization of TgVP1 and TgCPL in the same compartment was also confirmed by IEM (Fig. 3C and D), showing labelling of both the limiting membrane and internal vesicles (Fig. 3C and D, arrow in a vesicle in Fig. 3C).

The T. gondii aquaporin water channel, TgAQP1, has a high similarity to the plant aquaporins known as TIPs, which are found in lytic plant vacuoles (Pavlovic-Djuranovic et al., 2003; Zeuthen et al., 2006). In addition, a motif analysis of representative proteins from the major intrinsic protein (MIP) family of aquaporins supports the close relationship between TgAQP1 and plant γ-TIPs (see Fig. S2). Because the localization of TgAQP1 has not been reported, we prepared a polyclonal antibody against a C-terminal peptide (LGYVGTHAYHNPVPLRFLNFRGL) of the TgAQP1 gene. Antibody specificity of anti-TgAQP1 was compared by immunofluorescence and Western blot analyses in cells overexpressing an epitope-tagged version of TgAQP1 (Fig. S3). The anti-TgAQP1 antibody did not show a clear and defined reaction when used with wild-type parasites. This could be due to the low level of expression of the channel. In this regard, it should be mentioned that some pumps and transporters are found in very low numbers in small vesicles although they are functional. For example, it has been calculated that only one vacuolar H+-ATPase pump is present per synaptic vesicle (Takamori et al., 2006). However, there is transcript and proteomic information in the Toxoplasma database (ToxoDB) that indicates the expression of this gene in tachyzoites. TgAQP1 detection by antibodies against TgAQP1 and the epitope marker (c-myc) showed a similar discrete localization to a large vacuolar structure in extracellular tachyzoites (data not shown). This localization was consistently observed in a large number of cells (Figs 4 and S3B and C) from multiple IFA preparations. Further analysis using TgAQP1 overexpressing cells and the affinity-purified antibody for TgAQP1 showed that the T. gondii aquaporin colocalized with both TgCPL (Fig. 4B) and TgVP1 (Fig. 4C) at the site of a large vacuolar structure as visualized by DIC microscopy. Interestingly, the interaction between TgVP1 and TgAQP1 was sometimes complex and involved varying levels of localization to different vacuole compartments that were in physical contact (i.e. budding or fusing) (Fig. S3E–H and Movie S2). These instances suggest a dynamic relationship between the vacuoles where these markers are present (Fig. 4C and Movie S2, see smaller vacuoles labelled with both TgAQP1 and TgVP1). The affinity-purified antibody against TgAQP1 showed a reaction against a protein of the expected size of 27 kDa by Western blot analysis of T. gondii lysates of two clones overexpressing TgAQP1 (Fig. S3D, lanes 1 and 2). The clones overexpressing the TgAQP1 gene with a triple myc tag showed a reaction with the antibody at 32 kDa (Fig. S3D, lanes 3 and 4). No reaction was detected with preimmune serum (not shown).

Figure 4.

Localization of T. gondii aquaporin-1 (TgAQP1) in extracellular tachyzoites to a large vacuolar structure. To help ascertain the correct location of TgAQP1, tachyzoites overexpressing the native version of TgAQP1 were used. Detection of TgAQP1 was performed using an affinity-purified antibody against the C-terminal region of TgAQP1 (1:200).
A. Localization of TgAQP1 to a large vacuole, evident by DIC overlay and IFA (arrow).
B. Colocalization of TgAQP1 (green) with endogenous TgCPL (red). Arrow in merge shows colocalization (yellow) of TgAQP1 (green) with TgCPL (red).
C. Colocalization of TgAQP1 (green) with endogenous TgVP1 (red). Bottom arrow in merge shows a vacuole and localization of TgVP1 (guinea pig anti-serum at 1:100) where TgAQP1 is located. Top arrow shows colocalization of TgAQP1 and TgVP1 in a second vacuole or putative acidocalcisome. Scale bars: A = 5 µm, B = 3 µm and C = 5 µm.

Our data demonstrate the consistent presence of a large vacuolar compartment in T. gondii extracellular tachyzoites. This structure contains a hydrolytic cathepsin L-like protease (TgCPL) as well as a proton pump commonly found in plant vacuoles (TgVP1), thereby suggesting a lytic function to this organelle (see also Parussini et al., 2010). Furthermore, the limiting membrane of this structure contains an aquaporin water channel with specific similarities to the plant γ-TIP aquaporins (TgAQP1). We will refer to this organelle as the plant-like vacuole or PLV.

Interaction of the PLV with other vacuoles

Based on the electron microscopy in Fig. 2 it is evident that the PLV interacts with other vesicles, some of them appearing to be acidocalcisomes (see inset in Fig. 2B). The structural details of the two-dimensional organization of the PLV in these thin sections reveal its approximate size, preferential location in the cell and multi-vesicular nature. However, a general perspective of the organelle as a whole, its interaction with other cell compartments and its substructure at high resolution is better attained with the use of three-dimensional electron microscopy using electron tomography. This analysis revealed in more detail the multi-vesicular nature of the organelle, presenting vesicles of different sizes and varying levels of interaction with the PLV (i.e. vesicles inside, outside or in contact with the PLV membrane) (Fig. 2D–G and Movie S3). These results suggest that the PLV is a highly dynamic feature of the cellular architecture of extracellular tachyzoites.

Functional analysis of the PLV

Characterization of TgVP1 activity in PLV-enriched fractions. The presence of a vacuolar proton pyrophosphatase (TgVP1) indicates a significant role of the enzyme in the overall function of the organelle. In order to functionally characterize the biochemical role of this transporter, we generated a tachyzoite RH clone overexpressing the enzyme (see Experimental procedures) (TgVP1-OE). These cells were used, together with the wild-type cells (RH strain), to characterize the V-H+-PPase activity in PLV-enriched fractions (see below). TgVP1-OE had a higher content of TgVP1 as detected by enzymatic activity (225.08 ± 3.85 and 45.77 ± 4.51 nmol PPi hydrolysed min−1 × mg of protein in the fraction enriched in PLVs from overexpressing and control tachyzoites, respectively) (Table 1), and IFA (Fig. 1B). V-H+-PPase activity was assessed measuring aminomethylenediphosphonate (AMDP)-sensitive production of Pi from pyrophosphate (PPi) under appropriate conditions (Rodrigues et al., 2000). AMDP is a specific inhibitor of the V-H+-PPase (Zhen et al., 1994; Rodrigues et al., 2000; Drozdowicz et al., 2003) that we used to discriminate the TgVP1 activity from the activity of other pyrophosphatases. Because the hydrolysis of PPi releases energy that can be used to pump protons, the activity of TgVP1 could also be measured directly by measuring proton transport using acridine orange (Rodrigues et al., 2000). This proton transport activity was also increased in TgVP1-OE cells (compare slopes of graphs in Fig. 5A and B).

Table 1.  Enrichment of fraction 1 in AMDP-sensitive PPase activity.
EnzymeParasitesP3Fraction 1a
  • a. 

    Fraction 1 was obtained through subcellular fractionation of T. gondii tachyzoites following the steps described in the Fig. S4.

  • b. 

    This activity is the AMDP-sensitive pyrophosphate hydrolysis activity measured as explained under Experimental procedures.

  • c. 

    Protocols for the determination of these enzyme activities are explained under Experimental procedures.

  • Comparison of the values for columns P3 (pellet 3 from fractionation schematic in Fig. S4) and fraction 1 show a significant increase in the activities for the PLV markers (V-H+-PPase and acid phosphatase) and not in the activity of the soluble PPase (a cytosolic enzyme) in RH tachyzoites. The activity of the V-H+-PPase is higher in the fraction obtained from the TgPPase-OE (OE) cells than in the fraction from RH tachyzoites.

  • These results are representative of more than 10 fractionation experiments.

V-H+-PPaseb (nmol min−1 mg−1)RH2.27 ± 0.2445.77 ± 4.51
OE19.32 ± 0.17225.08 ± 3.85
Acid phosphatasec (nmol min−1 mg−1)RH32.60 ± 0.5555.25 ± 1.27
OE12.64 ± 0.12105.05 ± 2.60
Soluble PPasec (nmol min−1 mg−1)RH14.00 ± 0.7711.38 ± 0.10
OE14.27 ± 0.5037.83 ± 1.72
Figure 5.

PPi-driven proton transport in PLV-enriched fractions. Acridine orange accumulates in an acidic compartment and the quenching of the fluorescence correlates with the ΔpH (inside acidic). This accumulation is strictly dependent on the presence of PPi indicating a pyrophosphatase activity. The slope of the tracings correlates with the activity of the enzyme. This uptake is reverted upon addition of nigericin, a K+-/H+-ionophore because of protons being released from the vesicular compartment. PLV-enriched fraction (0.15 µg protein ml−1) was added to a buffer containing 130 mM KCl, 2 mM potassium phosphate, 2 mM MgCl2, 10 mM HEPES, pH 7.2 and 3 µM acridine orange. PPi (0.1 mM), ATP (0.5 mM) and nigericin (0.5 µM) were added where indicated.
A. Activity of fractions obtained from TgVP1-OE cells.
B. Activity of fractions obtained from RH cells.
C. Effect of buffer composition on the PPi-stimulated proton transport. The buffer described above was supplemented with either 130 mM KCl (trace a), 65 mM KCl and 130 mM sucrose (trace b), 130 mM NaCl (trace c) or 250 mM sucrose (trace d).
D. Inhibitory effect by AMDP (trace a: 0; trace b: 40 µM, trace c: 100 µM; trace d: 500 µM). F.U. indicates arbitrary fluorescence units.

An iodixanol gradient (Rohloff et al., 2004) was used to isolate a fraction enriched in PLVs (Fig. S4A). PLV enrichment was estimated by measuring the AMDP-sensitive V-H+-PPase activity, and by detection of TgCPL by Western blot analysis. As TgVP1 is not expressed at high levels, isolation of PLVs from wild-type RH parasites did not result in fractions with high proton transporting activity (Fig. 5B), prompting the use of TgVP1-OE tachyzoites to characterize the TgVP1 proton-translocating activity. Fractionation of TgVP1-OE cells resulted in fraction 1 having the highest specific activity (Table 1) and also the highest percent of activity of TgVP1 (Fig. S4B). In addition, TgCPL was also enriched in fractions 1–3 (Fig. S4B, bottom panel). The PLV-enriched fraction did not show enrichment in mitochondrial (cytochrome c reductase activity) or cytosolic (soluble pyrophosphatase activity) markers (Fig. S4C and Table 1), but did show enrichment in acid phosphatase activity, a marker for lysosomes in other cells (de Duve, 2005) (Fig. S4B). A gene (TGME49_004080) encoding a putative acid phosphatase is present in the T. gondii database, and peptides of the encoded protein were enriched in fraction 1 in our proteomic studies (Fig. S4, and S.N.J. Moreno and V. Carruthers, unpubl. results).

Proton transport was detected in the enriched PLV fractions from both TgVP1-OE and wild-type tachyzoites (Fig. 5) using acridine orange (Rohloff et al., 2004) indicating the presence of sealed vesicles. Acridine orange accumulates in acidic compartments and the quenching of its fluorescence correlates with the ΔpH (inside acidic). This accumulation is strictly dependent on the presence of PPi because there was no acridine orange uptake detected in the absence of PPi (Fig. 5A, trace -PPi) indicating a pyrophosphatase activity. The slope of the tracings correlates with the activity of the enzyme as more protons are actively pumped inside of sealed vesicles. Upon addition of nigericin, a K+-/H+-ionophore, protons are released from the vesicular compartment in exchange for potassium (see scheme in inset of Fig. 5). Further characterization of PPi-dependent proton uptake was conducted using different buffers to simulate varying environmental conditions relevant to tachyzoites as well as potential inhibitors of the PPase activity (Figs 5C and D and S5). Potassium had a stimulatory effect with the maximum proton transport observed at 130 mM KCl (Fig. 5C, trace a). Decreasing the amount of KCl from 130 to 65 mM reduced proton pumping activity relative to controls (Fig. 5C, trace b), as did replacing KCl with NaCl (Fig. 5C, trace c), or replacing KCl and NaCl, with sucrose (Fig. 5C, trace d). These results indicate that the TgVP1 is stimulated by K+, similar to V-H+-PPases from plant and other protists (Rea and Poole, 1986; Scott et al., 1998; Rodrigues et al., 1999).

Pyrophosphate-induced acidification rate was inhibited in a concentration-dependent manner by the pyrophosphate analogues AMDP and IDP (Figs 5D and S5A, respectively) and by high concentrations of either NaF or KF (Fig. S5C and D respectively). Chloroquine, a drug that alkalinizes acidic compartments, was also able to inhibit uptake or stimulate acridine orange efflux depending on the concentration used (Fig. S5B).

It was observed that PLV fractions from both the TgVP1-OE (Fig. 6C) cells and the wild-type tachyzoites (Fig. 6B) had similar ATP-driven acridine orange transport activity (compare Fig. 6B and trace b from Fig. 6C). However, PLV fractions obtained from TgVP1-OE tachyzoites had much higher PPi-driven acridine orange uptake (compare the slopes of the tracings of TgVP1-OE and wild-type tachyzoite fractions in Fig. 5A and B respectively). The stimulation of proton uptake by ATP indicates the presence of a H+-ATPase activity that was completely inhibited by 1 µM bafilomycin A1 (Fig. 6C, trace a), a specific inhibitor of V-H+-ATPases when used at low concentrations (Bowman et al., 1988). This ATP-driven acridine orange uptake was further stimulated by addition of PPi, at a rate faster than that obtained with ATP alone (Fig. 6C, trace b). When the order of additions was reversed, PPi caused fast acridine orange uptake but addition of ATP did not lead to further accumulation of the dye (Fig. 6C, trace c). In both cases, acridine orange was immediately released by addition of nigericin but no additive effects were observed suggesting that the ATP-driven and PPi-driven proton pumps are located in the same compartment (i.e. the PLV) in this fraction.

Figure 6.

Localization of V-H+-ATPase in the PLV.
A. Immunogold electron microscopy showing labelling of the large vacuole and internal vesicles with a rabbit polyclonal antibody against D. discoideum V-H+-ATPase (Moreno et al., 1998) (10 nm gold particles). Bar = 0.1 µm.
B. ATP-driven H+ transport in PLV-enriched fractions from wild-type parasites.
C. Proton transport in PLV-enriched fractions from TgVP1-OE parasites stimulated by ATP (traces a and b) in the presence of bafilomycin (trace a). Addition of PPi after ATP shows a significant increase in proton transport. Trace c shows first the transport stimulated by PPi, and addition of ATP does not affect the rate when added subsequently. All incubations were done with PLV fraction (0.15 µg protein ml−1) in the same buffer described in the legend for Fig. 5A and ATP (1 mM), PPi (0.1 mM) and nigericin (0.5 µM) were added where indicated. F.U. indicates arbitrary fluorescence units.

Based on the biochemical evidence of ATP-driven proton transport in T. gondii PLV fractions (Fig. 6B), we investigated the localization of the V-H+-ATPase using antibodies against the purified V-H+-ATPase from Dictyostelium discoideum that we previously used to detect localization of this enzyme in T. gondii (Moreno et al., 1998). These antibodies react with several subunits of T. gondii V-H+-ATPase, as detected by Western blot analysis (Moreno et al., 1998). Interestingly, by IFA these antibodies labelled weakly the plasma membrane and very strongly an intracellular vacuole that at that time we were unable to identify (Moreno et al., 1998). IEM revealed labelling of the membrane of the PLV and of internal vesicles (Fig. 6A), a labelling very similar to that obtained with antibodies against TgVP1 (Fig. 3C) and TgCPL (Fig. 3D).

Ion homeostasis and salt stress.  Further experiments on enriched PLV fractions showed the presence of additional ion exchange mechanisms. Addition of 80 mM NaCl (Fig. 7A, trace b) but not of 80 mM KCl (Fig. 7A, trace c) after PPi-driven acridine orange uptake reached a steady-state level, resulted in acridine orange efflux, suggesting the activity of a Na+/H+ exchanger in the PLV fraction (see scheme in inset of Fig. 7A).

Figure 7.

TgVP1-overexpressing cells are more resistant to salt stress and evidence for a Na+/H+ exchanger.
A. Upon addition of sodium, acridine orange is released from the vacuolar compartment because of alkalynization. This is because there is an exchange of sodium for protons strongly suggesting the operation of a sodium proton exchanger (see scheme). Effect of 80 mM NaCl (trace b) or 80 mM KCl (trace c) addition. F.U., arbitrary fluorescence units at 470–526 nm. Other conditions as in Fig. 5A.
B. Plaque assays comparing growth of RH versus TgVP1-OE cells pre-incubated for 30 min in invasion media (IM) or PBS containing additional 150 mM NaCl (NaCl) (287 mM in total).
C. Plaque number was quantified in four independent wells for each condition. Bars represent standard errors.

We exposed extracellular tachyzoites to high concentrations of NaCl (287 mM final concentration) during short periods of time (30 min), allowed them to infect host cells and evaluated their capacity to form plaques in a fibroblast culture at 9 days after infection. Relative to the number of plaques formed using invasion media (IM), RH cells showed 90% inhibition in the number of plaques formed in the high salt condition (Fig. 7B and C). However, tachyzoites overexpressing the V-H+-PPase (TgVP1) were significantly more resistant to this treatment, and the number of plaques was only inhibited by 56% under identical treatment (Fig. 7B and C). The number of plaques indicates the number of cells that survive the stress treatment, and the results show that TgVP1-OE cells are more resistant to the salt stress. TgVP1 may be facilitating the sequestration of sodium into the PLV and helping the parasite to survive under these extreme conditions. As shown in Fig. 7A sodium transport into the PLV could be coupled to proton transport by the TgVP1.

Calcium storage.  Addition of 100 µM CaCl2 to PLV-enriched fractions resulted in acridine orange efflux (Fig. 8A, trace a), because of calcium being transported into a compartment in exchange for protons (see scheme in inset of Fig. 8A). This result suggests the presence of a Ca2+/H+ exchanger and indicates that the organelle could be, as is the case in the plant vacuole or the mammalian lysosome, an important acidic calcium store (Patel and Docampo, 2010). To test this, we used glycyl-L-phenylalanine-naphthylamide (GPN), which is specifically hydrolysed in the lysosome of a variety of different cell types by a cathepsin C protease. This results in an increase in osmolarity within the lysosome leading to its swelling and release of stored calcium into the cytosol (Haller et al., 1996; Christensen et al., 2002; Lloyd-Evans et al., 2008). In this regard, our proteomic analysis of the PLV-enriched fractions (S.N.J. Moreno and V. Carruthers, unpubl. results) revealed the presence of two cathepsin C proteases. In agreement with our hypothesis, addition of GPN to fura-2AM-loaded tachyzoites in the nominal absence of extracellular Ca2+ (1 mM extracellular EGTA added) resulted in Ca2+ release to the cytosol, which was independent of Ca2+ release induced by thapsigargin, a drug that releases Ca2+ from the endoplasmic reticulum (Fig. 8C, trace a, two independent peaks are observed, evidence of two separate pools) (Moreno and Zhong, 1996). Similar results were observed when the order of additions was reversed (Fig. 8C, trace b). GPN also released calcium after nigericin (Fig. 8B, trace b) indicating the presence of a distinct acidic calcium pool (acidocalcisomes), which does not contain a cysteine protease. Similar results were obtained when GPN was added first and nigericin was added second (Fig. 8B, trace a).

Figure 8.

The PLV contains calcium, which can be released by GPN.
A. Biochemical evidence of a calcium/proton exchange mechanism in tachyzoites. Upon addition of calcium (100 mM, trace a) to a previously PPi-energized PLV fraction, a discernable release of protons is evident that was not observed in the control experiment (see scheme for proposed scenario). Other conditions as in Fig. 5A.
B. and C. Effect of GPN on intracellular calcium levels. T. gondii tachyzoites were loaded with fura-2AM and cytosolic calcium concentration was monitored in BAG containing 1 mM EGTA at a 2 × 107 cells ml−1 final concentration in a cuvette and scanned in a Hitachi F-4500 spectrofluorometer.
B. Cytosolic calcium concentration upon sequential addition of GPN followed by nigericin (NIG) (trace a). Trace b shows a similar response when the order of addition is reversed with GPN addition after the addition of NIG. Baseline calcium concentration with no additions is shown in trace c.
C. Cytosolic calcium concentration upon sequential addition of GPN and thapsigargin (TG) (trace a). Trace b shows a similar response when the order of addition is reversed with TG addition followed by GPN. Final concentration of agonists were: GPN, 20 µM; thapsigargin, 2 µM; nigericin, 1 µM.

Response to environmental stress.  HgCl2 is a specific inhibitor of aquaporin water channels (Niemietz and Tyerman, 2002). With the aim of testing the physiological function of the aquaporins, the ability to tolerate toxic levels of HgCl2 was investigated in wild-type (RH) and TgAQP1 overexpressing tachyzoites. When RH tachyzoites were incubated in 1 µM HgCl2 for 5 min, the cells became rounded and the PLV occupied a larger proportion of the cellular volume relative to untreated control cells (Fig. 9A, RH). In contrast, tachyzoites overexpressing TgAQP1 showed a much greater tolerance to this stress. Overexpressing mutants of TgAQP1 were less rounded, maintained their typical size and volume, and the PLV retained its usual shape and size when exposed to 1 µM HgCl2 (Fig. 9A, TgAQP-OE, compare no treatment with mercury treatment). We performed an analysis of overall shape by quantifying the circularity of extracellular tachyzoites when in the presence or absence of 1 µM HgCl2 (Fig. 9B). The circularity of RH cells increased significantly (i.e. the cells became more rounded), the circularity index changing from 0.63 ± 0.07 (SD) to 0.81 ± 0.09 (SD) when incubated in 1 µM HgCl2 (Fig. 9B). The TgAQP1 overexpressing cells had an average circularity similar to the control values of RH at 0.62 ± 0.09 (SD). However, when exposed to 1 µM HgCl2 the general shape of the parasites remained unchanged with an average circularity of 0.63 ± 0.09 (SD).

Figure 9.

Response to HgCl2 in TgAQP1 overexpressing extracellular tachyzoites. Extracellular tachyzoites were collected, removed from culture media and resuspended in buffer A with glucose (BAG).
A. RH and TgAQP1 overexpressing tachyzoites were either kept in BAG for 5 min (‘no treatment’) or incubated in BAG with HgCl2 at a final concentration of 1 µM for 5 min (‘mercury treatment’). Cells were then fixed with paraformaldehyde for 1 h and mounted on coverslips for visualization using DIC microscopy. Scale bars for A = 10 µm.
B. Morphometric analysis of RH and TgAQP1 overexpressing cells in the absence (black bars) or presence (grey bars) of 1 µM HgCl2. Circularity measurements were made on 50 randomly chosen parasites for each treatment. Error bars are standard deviation, ***P < 0.001.

The PLV in intracellular parasites

During intracellular replication of T. gondii, the localization of TgCPL and TgVP1 changed as both took on a punctate distribution and their expression no longer localized to a highly visible (i.e. by DIC microscopy) vacuolar structure (Fig. 10A). Proteomic studies of the PLV-enriched fractions (S.N.J. Moreno and V. Carruthers, unpublished) resulted in the identification of several Rab (Ras-related proteins in the brain) proteins, among them Rab7. Rab proteins are members of the wider Ras superfamily of GTPases and are essential regulators of membrane trafficking. Rab7 is usually associated with late endosomes and lysosomes where it regulates membrane fusion (Stenmark and Olkkonen, 2001). We overexpressed an haemaglutinin (HA) epitope-tagged TgRab7 in T. gondii tachyzoites and found that it localizes to a compartment that contains vesicles that label with anti-TgVP1 in intracellular parasites (Fig. 10B). TgRab7 was specifically observed in these vacuoles while TgVP1 was also observed in other parts of the cells (acidocalcisomes and/or other vacuolar compartments). During intracellular replication the circular profile of the PLV became diminished and the majority of the TgVP1 signal was localized in a post-Golgi compartment as shown by the relative location of the Golgi marker TgGRASP55 (Fig. 10C). Under our experimental conditions, we did not observe colocalization between anti-TgCPL and TgRab7 labelled with anti-HA (not shown). These results suggest that intracellular parasites may not have PLVs and the compartment labelled with Rab7 and TgVP1 corresponds to the previously described ‘VP1 compartment’ (Harper et al., 2006) that would be a pre-PLV compartment with characteristics of a late endosome. This compartment may then form part of the PLV during the transition from intracellular to extracellular parasites. For instance, parasites in one-cell parasitophorous vacuoles retain a visible PLV showing clear localization of TgVP1, TgCPL and TgAQP1 (Figs 10D and S6, one parasite panels). However, from the two-cell stage onward, the structure of the PLV is no longer noticeable by DIC or antibody labelling of these markers. This fragmentation (Larson et al., 2009) and apparent loss of the PLV result in anti-TgVP1 and anti-TgAQP1 only partially colocalizing with anti-TgCPL (Fig. 10A). Parussini et al. (2010) describe in detail in the accompanying article the PLV fragmentation during daughter cell formation.

Figure 10.

The PLV fragments after host cell invasion.
A. Lack of colocalization of anti-TgCPL used at a dilution of 1:400 (green) and anti-TgVP1 used at a dilution of 1:4000 (red). The arrows indicate a typical appearance of vesicles labelled with anti-TgVP1 apical to the nucleus. h-Tert monolayers were infected with 1.5 × 106 tachyzoites and fixed 20 h later for IFA assay.
B. Colocalization of anti-TgVP1 (green) and antibodies against the HA tag of TgRab7 used at a dilution of 1:800 (red). The arrows points at vesicles labelled with the anti-TgVP1 antibody.
C. Labelling by anti-TgVP1 is in a post-Golgi compartment. Parasites were transfected with the plasmid pTubGRASP55-RFP/sagCAT. The red labelling comes from the direct fluorescence of the red fluorescent protein (arrow points at the red-labelled compartment in left panel). The labelling of vesicles with anti-TgVP1 is indicated by arrows in the TgVP1 and Merge panels.
D. Replicating, intracellular parasites were labelled at different stages of division with anti-TgVP1 antibodies. The arrows show the appearance of the PLV as the cells divide. Anti-TgVP1 primary antibody was used at 1:4000, and goat anti-rabbit Alexa 488 as secondary antibody was at 1:1000. Blue fluorescence = DAPI. Scale bars = 5 µm.

The PLV appears in recently egressed extracellular parasites

Because of the observed fragmentation of the PLV during intracellular replication, further analysis was conducted on extracellular tachyzoites to better understand its occurrence during the entire extracellular stage. Intracellular tachyzoites from semisynchronized cultures were released by scrapping off the host monolayer and passage through a syringe needle. These parasites were immediately collected and incubated under two different conditions (BAG vs. IM, see description under Experimental procedures and legend to Fig. 11) for 0, 2 and 4 h, during which time the shape of the PLV (closed vs. open) and number of PLV-containing parasites were evaluated (Table 2). Figure 11 shows the results obtained from one representative experiment. It is noticeable that under the conditions of the extracellular buffer BAG, the parasites show a more open PLV when compared with the same parasites in IM (Table 2). Analysis of invading tachyzoites showed that the PLV was localized to the apical end of parasites at the onset of invasion (Fig. 11B invading tachyzoite). In this regard, a similar clear vacuole appears in invading tachyzoites in Fig. 5C and accompanying movie of Lovett and Sibley (2003). As described above, once inside the host cell the PLV diminished as replication progressed. The presence of an open (i.e. large and circular) PLV in recently released as well as actively invading parasites indicate that the PLV is a normal aspect of tachyzoite morphology and not an artefact observed in degenerating or aging cells.

Figure 11.

The PLV is evident in extracellular parasites soon after release from host cells.
A. Intracellular tachyzoites were obtained after scrapping off infected host monolayers and fixed immediately (time 0), or incubated for 2 and 4 h in BAG or IM. Anti-TgCPL and anti-TgVP1 antibodies were used to visualize the PLVs at those times. The most common phenotype is shown. Table 2 shows the quantification of these cells.
B. Tachyzoites were allowed to invade h-Tert monolayers for 5 min at 37°C after a pre-incubation of 15 min on ice. Cells were immediately fixed, and anti-TgCPL and anti-TgVP1 antibodies were used to visualize the PLV. Antibody concentrations are indicated in the legend for Fig. 3. Scale Bars = 5 µm.

Table 2.  Number of open versus closed PLVs in recently released tachyzoites.
ConditionsTotal number of parasitesaNumber of open vacuolesbNumber of closed vacuoles
  • a. 

    These numbers are the average number of parasites counted in five different fields. Parasites with visible vacuoles by DIC and labelled with anti-TgVP1 were counted.

  • b. 

    Numbers given in parentheses represent the percentage of cells in open or closed configuration. Cells with no clear PLV were not counted as either open or closed, and represented less than 25% of the total population.

  • The results presented are from one experiment out of three. The other two experiments gave similar results (total cells evaluated were 189 in BAG and 177 in IM).

  • Tachyzoites were collected by scrapping off the host monolayer and releasing parasites by passing them through a 25 G syringe needle. Cells were collected in either BAG or IM, and the numbers in the table are the ones obtained immediately after isolating the parasites (Time 0 in Fig. 11). Vacuoles were considered open when their diameter was larger than 0.5 µm. Vacuoles were considered closed when their diameter were equal or smaller than 0.5 µm.

BAG3826 (68%)3 (7%)
IM357 (24%)23 (67%)

Discussion

We report the identification of a previously uncharacterized organelle in tachyzoites of T. gondii. This organelle is a large multi-vesicular structure that possesses similarities to the central vacuoles found in plant cells. Similarities in structure, composition and potential functions prompted us to name this organelle the plant-like vacuole, or PLV. This organelle labels with antibodies against proteins with great similarity to vacuolar plant pumps and channels, such as a K+-sensitive V-H+-PPase (TgVP1), and an aquaporin water channel (TgAQP1). It is interesting to note that Drozdowicz et al. (2003) detected the localization of TgVP1 to a large vacuole in invading tachyzotes that we can now identify as the PLV. Sequence analysis has shown that TgAQP1 groups with other coccidian apicomplexan aquaporins such as Emeria sp. (Pavlovic-Djuranovic et al., 2003). Interestingly these aquaporins share a plant-like divergence from other aquaporins (e.g. Plasmodium sp.) in that the highly conserved arginine residue at the aromatic arginine region of the channel pore (Ar/R region) has been replaced with a valine residue (Pavlovic-Djuranovic et al., 2003). This substitution is characteristic of TIPs from plants and has a significant function in determining transport capabilities of these channels. Such analyses have been used to infer that apicomplexan aquaporins may have been adopted from different taxonomic sources, with the Plasmodium sp. aquaporins deriving from bacterial glycerol facilitators and the T. gondii aquaporin originating from plant TIPs (Pavlovic-Djuranovic et al., 2003). This analysis further supports the plant-like nature of TgAQP1 and of the PLV. In addition, all plant vacuoles contain the V-H+-PPase supporting the similarity of the PLV to plant vacuoles. Physiological evidence revealed further similarities to plant vacuoles such as the presence of a V-H+-ATPase, a Na+/H+ exchanger and a Ca2+/H+ exchanger, and the storage of calcium (Maeshima, 2001; Becker, 2007).

Many proteases reside in the lumen of plant vacuoles and are used for protein degradation to produce amino acids that are recycled for metabolic processes that take place outside the vacuole (Muntz, 2007). Cathepsin L-like proteases (Bethke et al., 1996; Vincent and Brewin, 2000) have been described in a number of plant vacuoles (Boller and Kende, 1979) and our results show the localization of a cathepsin L-like protease (TgCPL) in the PLV. The observation that TgCPL is inside vesicles as well as in the membrane of the PLV is interesting. It is possible that this lytic enzyme has two distinct locations based on the maturation of its propeptide or that it is temporarily compartmentalized to avoid contact with its substrates as is the case in the plant vacuole (Jiang et al., 2001).

Our results show that the PLV is a prominent feature of extracellular tachyzoites and with multiple potential functions as is the case for the plant vacuole in the plant cell. We show evidence for its potential role in the transport of sodium and calcium, calcium storage, and in resistance to environmental stresses, although we cannot rule out that some of the functions attributed to the presence of TgAQP1 and TgVP1 in the PLV could be shared with other organelles in which these proteins are located. Transgenic plants overexpressing the V-H+-PPase are much more resistant to high concentrations of NaCl than the wild-type strains (Gaxiola et al., 2001). When T. gondii tachyzoites egress from the host cell they are exposed to an abrupt change in sodium from the intracellular concentration of 2–5 mM to the extracellular concentration of 100–150 mM. Parasites have to survive under these conditions long enough to find other host cells to invade. Our salt stress experiments combined with Na+/H+ exchange activity observed in PLV fractions, strongly suggest that TgVP1 might have a role in the homeostasis of intracellular sodium by helping to sequester it into the PLV.

Our analysis of semisynchronized cells shows that the PLV is already present in freshly egressed tachyzoites as well as invading tachyzoites, indicating that the PLV fulfils important physiological functions and is not a consequence of aging parasites. We find a significant difference in the appearance and size of the PLV that is related to the conditions of the extracellular media. Extracellular tachyzoites incubated in basic extracellular media (BAG) show a more prominently open PLV. Parasites collected in Ringer buffer also gave a similar result (not shown). When parasites were prepared and incubated in a media containing significantly less NaCl (IM, 50% less NaCl) the appearance of the PLV was diminished and collapsed. It is possible that ionic composition plays a large role in the morphology of the PLV with NaCl being at least one important determinant of its state. Future work analysing in more detail the specific component(s) that produces these changes in the morphology of the PLV will help clarify this phenomenon.

During intracellular replication, the colocalization of TgVP1 and TgCPL is not clearly evident or consistent. Instead, TgVP1 colocalizes with TgRab7 to a vacuolar compartment, potentially a late endosome. This is interesting because it might shed light into the generation of this organelle as the parasites leave the host cell and prepare to confront the external environment. It is possible that the PLV is formed through the fusion of a late endosome (because of its labelling with Rab7) containing TgVP1 [the ‘VP1 compartment’ described by (Harper et al., 2006)] and TgAQP1 with lysosome-like vacuoles containing TgCPL. This structure is then maintained in extracellular tachyzoites until after invasion of a new host cell (see our proposed model in Fig. S7). We propose that this VP1 compartment functions as a pre-PLV compartment in intracellular parasites, which fuses with other vesicles to form the PLV in extracellular parasites.

In this work we have also characterized for the first time the proton pumping activity of T. gondii V-H+-PPase (TgVP1). Pyrophosphate-driven proton transport was inhibited by the PPi analogues IDP and AMDP and was stimulated by K+ ions. In this regard, AMDP and other PPi analogues have been shown to inhibit T. gondii growth (Rodrigues et al., 2000; Drozdowicz et al., 2003). In plants, the V-H+-PPase and the V-H+-ATPase are located in the same membrane (Rea et al., 1992) and in our enriched PLV fractions, we observed that there is no additive acidification induced by ATP or PPi suggesting that both pumps are present in the same vacuole.

Plant lytic vacuoles, which are acidic and rich in hydrolases, are considered as equivalent to the animal lysosome and are recognized by the presence of γ-TIP (Becker, 2007). Sequence analysis of TgAQP1 demonstrated its similarity with plant γ-TIPs (Fig. S2), which are specific of the plant lytic vacuole (Martinoia et al., 2007). We used the major intrinsic protein database (MIPDB) as a reference to retrieve representative protein sequences from each MIP subfamily (there are ∼8 MIP subfamilies, plants contain three subfamilies: PIPs, TIPs and NIPs). We identified a total of five conserved motif regions in the TgAQP1 sequence (Fig. S2A and B). Of these five motifs, four were exclusively associated with the plant γ-TIP family (Fig. S2A, motifs 1–4, green). The fifth motif belonged to the prokaryotic glycerol facilitator subfamily and is present in several taxonomically divergent taxa (e.g. Plasmodium sp., Neospora sp. and plants). These results confirm previous sequence analysis showing that TgAQP1 has a high similarity to plant γ-TIPs (Pavlovic-Djuranovic et al., 2003; Zeuthen et al., 2006) and provide further support to the similarity between the T. gondii vacuole and the plant lytic vacuole.

Plant cell vacuoles are part of the eukaryotic endomembrane system and exhibit a wide diversity in form and function (Marty, 1999; Becker, 2007). Plants have evolved a large central vacuole to allow them to increase their cell volume without the need to invest in cytoplasm and other organelles (Becker, 2007). Many protists, especially those that live in water, posses contractile vacuoles involved in osmoregulation as well as other acidic vacuoles like acidocalcisomes for storage of phosphorus and cations. T. gondii contains acidocalcisomes, and it is possible that it acquired an organelle similar to the one found in plants through convergent evolution. Plant vacuoles are thought to have a division of labour between storage and lytic vacuoles, and a similar situation might be occurring in T. gondii tachyzoites. Our characterization of the PLV has also demonstrated that it not only has a high degree of physical interaction with acidocalcisomes but also possesses markers that are common to acidocalcisomes (e.g. TgVP1 and H+-ATPase). It may be that acidocalcisomes in T. gondii have an analogous role as plant storage vacuoles and that the PLV is fulfilling functions in T. gondii tachyzoites that are done by lytic vacuoles in plants.

The PLV is a proton sequestering organelle driven by PPi and ATP and appears to possess several other ion exchange mechanisms. A model showing how these pumps would work in combination with other transporters (e.g. sodium, calcium and other cations) is shown in Fig. 12. Considering the role of the plant vacuole, and based on our findings, the PLV could have an homeostatic function providing protective mechanisms needed by extracellular tachyzoites to survive in diverse biological environments. Although T. gondii is an obligate intracellular parasite, it is still not completely understood how tachyzoites disperse throughout the host organism as they can cross the intestinal epithelium, disseminate into the deep tissues and actively traverse biological barriers such as the placenta and the blood–brain barriers (Barragan and Sibley, 2003; Tardieux and Menard, 2008). We propose that the PLV plays a critical role in this biological context. The PLV could also provide ‘turgor pressure’ to the parasite to facilitate active egress from the host cell or active penetration into a new host cell in the same way that the plant vacuole provides turgidity. During intracellular replication the pre-PLV compartment may play additional roles involved in the endocytic/exocytic pathways where the sorting of proteins targeted to various organelles necessary for invasion and egress occurs. The accompanying manuscript by Parussini et al. (2010) provides evidence for a role of this compartment in the proteolytic maturation of pro-proteins targeted to micronemes. This plant-like organelle has not been described before in any non-plant organism. Its detailed characterization will help in better understanding the adaptive responses of T. gondii tachyzoites during the extracellular stage when it has the capability of invading numerous types of host cells. It is quite likely that the PLV is also present in other apicomplexan parasites and further studies will be relevant to the identification of novel targets for chemotherapy against T. gondii and other parasites as well.

Figure 12.

Schematic representation of the plant-like vacuole. The H+ gradient is established by a vacuolar H+-ATPase (V-H+-ATPase) and a vacuolar proton pyrophosphatase (V-H+-PPase). An aquaporin channel would transport water or other osmolytes that could help the parasite deal with environmental stress. Other potential transporters include Na+/H+ and Ca2+/H+ exchangers. As in the plant vacuole, other transporters may be present, which could transport other cations using the proton gradient generated by the proton pumps. Some internal vesicles and acidocalcisomes are also shown.

Experimental procedures

Parasites

Toxoplasma gondii tachyzoites (RH strain) were grown in h-Tert human fibroblasts (Farwell et al., 2000) as described before (Moreno and Zhong, 1996). These cells grow in DMEM media containing 1% FBS.

For semisynchronization of cultures, h-Tert cells cultured in 75 cm2 flasks, were infected with 3.7 × 107 tachyzoites/flask for 2 h, extracellular parasites thoroughly washed and the cultures allowed to grow for 35–40 h. At this time, extracellular parasites were removed by washing with fresh IM (DMEM containing 20 mM HEPES pH 7.4 with 1% FBS) three times and the cultures allowed growth for two more hours in IM. Subsequently, the extracellular tachyzoites were washed off and the intracellular tachyzoites collected in fresh IM by scrapping off the host monolayer and purifying the parasites by filtration through a nucleopore membrane. The isolated tachyzoites were centrifuged and resuspended in IM without serum or buffer A plus glucose (BAG) (116 mM NaCl, 5.4 mM KCl, 0.8 mM MgSO4, 50 mM HEPES, pH 7.2, 5.5 mM glucose) at a concentration of 5 × 107 tachyzoites ml−1, and incubated for 2 or 4 h at 37°C.

For stress experiments involving mercury tolerance, tachyzoites were incubated in BAG with 1 µM HgCl2 for 5 min, fixed with 4% paraformaldehyde and mounted on coverslips. Using Image J software (NIH), circularity measurements (a metric of roundness where 1 equals a perfect circle) were made on 50 randomly chosen cells from each treatment to determine changes in overall cell shape when in the presence or absence of 1 µM HgCl2.

For salt stress experiments freshly egressed tachyzoites were purified and washed in IM, incubated for 15 and 30 min under stress conditions (described in the legend for Fig. 7) and subsequently added to the regular culture medium. For plaque assays confluent monolayers of fibroblasts grown in six-well plates were infected, in triplicate, with 225 tachyzoites per well. The parasites were allowed to plaque for 9 days, fixed and stained as described (Roos et al., 1994).

Isolation of PLV-enriched fractions

Isolation of PLV-enriched fractions was according to a modification of a method described previously for the isolation of contractile vacuoles from Trypanosoma cruzi (Rohloff et al., 2004) (Fig. S4). Tachyzoites (∼1–2 × 1010 cells) were purified and washed with lysis buffer (125 mM sucrose, 50 mM KCl, 4 mM MgCl2, 0.5 mM EDTA, 20 mM K-HEPES pH 7.2, 5 mM dithiothreitol, protease inhibitors (0.2% v/v), 12 mg ml−1 DNAse, 12 mg ml−1 RNAse and 8 mg ml−1 nocodazole). The pellet was mixed with 2× wet weight silicon carbide and grinded for 60 s. This mixture was resuspended in approximately 40 ml of lysis buffer and the suspension decanted and clarified by three low speed centrifugations. The supernatant was centrifuged at 15 000 g for 10 min, and the new supernatant centrifuged at 100 000 g for 60 min. The resulting pellet (P3) was homogenized and loaded into the 20% layer of an iodixanol gradient containing 4 ml steps of 15%, 20%, 25%, 30%, 34% and 38%. The gradient was centrifuged at 50 000 g for 60 min. Fifteen fractions of 1.8 ml each were collected from the top of the gradient. Fractions 1–3 were the PLV-enriched fractions (Fig. S4).

Antibody generation and purification

Guinea pig and rabbit polyclonal antisera were raised against synthetic peptides corresponding to amino acids GLGPEVRSRTDALDA (between transmembrane region 11 and 12) and SGKNEYGMSEDDPRN (between transmembrane regions 6 and 7) of the TgVP1 sequence, respectively, and affinity purified by Covance Research Products (Berkeley, CA). The antibodies were shown to react with a protein of 72 kDa in subcellular fractions of T. gondii (Fig. S1C). No reaction was detected with preimmune serum for either antibody. A peptide (LGYVGTHAYHNPVPLRFLNFRGL) from the C-terminal domain of the TgAQP1 sequence was synthesized by Covance and used to generate antibodies in rabbit. The affinity-purified antibody showed a reaction against a protein of approximately 27 kDa (Fig. S3D).

Enzyme assays and calcium measurements

Pyrophosphatase activities were assayed by measuring phosphate release using the malachite green assay (Lanzetta et al., 1979). AMDP was used to distinguish between the vacuolar and the soluble activity. Acid phosphatase was assayed by measuring phosphate release from p-nitrophenylphosphate (Rodrigues et al., 2002) in acetate buffer pH 5.0.

PPi- and ATP-driven proton transport in PLV-enriched fractions was measured by changes in the fluorescence of acridine orange at excitation and emission wavelengths of 470 and 526 nm, respectively, using a Molecular Devices Microplate Reader. Fractions were incubated in a 200 µl final volume of a solution that is described in the figure legends, plus 3 µM acridine orange for 3 min prior to the addition of 100 µM PPi or 0.5 mM ATP. Nigericin was used to collapse the membrane potential generated. Each experiment was repeated at least three times with different fractionations, and the figures show representative experiments.

Intracellular calcium measurements were performed using fura-2AM (Invitrogen) as previously described (Moreno and Zhong, 1996).

Overexpression of the TgVP1, TgRab7 and TgAQP1

We isolated tachyzoites that overexpress the vacuolar proton pyrophosphatase by transfecting cells of the RH strain with the plasmid ptubTgVP1-FLAG/sag-CAT (Striepen et al., 1998), which contains the entire coding sequence of the TgVP1 gene (Drozdowicz et al., 2003) (GenBank: AAK38077.1; Toxodb: TGME49_048670). This plasmid was made by replacing the P30 gene with the coding sequence of the TgVP1 in the vector ptubP30-FLAG/sag-CAT (Luo et al., 2005). Selection was done in the presence of chloramphenicol, and one clone was selected for further analysis. TgVP1-OE had a higher content of TgVP1 as detected by Western blot (Fig. S1C), proton transport activity (Table 1) and immunofluorescence analyses (Figs 1A and B and 3A and B).

The entire open reading frame of TgAQP1 (GenBank: AAP33053.1; Toxodb: TGME49_015450) was amplified by RT-PCR from T. gondii cDNA. The forward primer (5′-AGATCTATGGACCAATTTGTTTTTTCAGGAGGTTC-3′) included the BglII restriction site (underlined), which is compatible with the BamHI site in the vector. The reverse primer (5′-CCTAGG GAGCCCCCTGAAGTTCAAGA-3′) omitted the stop codon and introduced an AvrII restriction site (underlined). The PCR product was cloned into the expression vector pDTM3 (provided by Dr Boris Striepen) between the BglII/AvrII sites, creating a triple-c myc epitope-tagged version of TgAQP1. The created pDTM3AQP-1 recombinant plasmid was confirmed by DNA sequencing. T. gondii RH tachyzoites were transfected with 50 µg of plasmid DNA and selected with 1 µM of pyrimethamine. An alternative construct was also made in the same vector but the stop codon of TgAQP1 was not omitted, therefore removing the triple-c myc epitope. For these constructs, AQP1 was detected with the peptide antibody to the C-terminal domain of TgAQP1 (see above).

The entire open reading frame of TgRab7 (GenBank: XP_002367234.1; ToxoDB: TGME49_048880) was amplified by RT-PCR from T. gondii cDNA. The forward primer (5′-TCCCCCGGGATGCCGC CCAAGAAGAAGG-3′) included the XmaI restriction site (underlined). The reverse primer (5′-CGCG GACGTCTCAGCAGCAGCCGCCG-3′) included the stop codon and introduced an AatII restriction site (underlined). The recombinant TgRab7 was cloned into the pCTH expression vector (provided by Dr Boris Striepen) between the XmaI/AatII sites. The pCTH-TgRab7 recombinant plasmid was confirmed by DNA sequencing. Twenty-five micrograms of recombinant plasmid (pCTH-TgRab7) were used to transfect T. gondii tachyzoites, which were inoculated in 12-well plates containing h-Tert fibroblasts grown in coverslips. Parasites were cultured for 16–20 h and used for IFA analysis.

Fluorescence microscopy

Toxoplasma gondii tachyzoites were harvested and washed with BAG or Ringer buffer (155 mM NaCl, 3 mM KCl, 1 mM MgCl2, 3 mM NaH2PO4-H2O, 10 mM HEPES, pH 7.2, 10 mM glucose) and fixed with 4% formaldehyde for 1 h. Immunofluorescence assays were performed as described (Luo et al., 2001) by using primary antibodies at the concentrations indicated in the legends. Secondary antibodies were Alexa 488 and Alexa 546 conjugated anti-rabbit IgG or Alexa 488 and Alexa 568 (1:500 or 1:1000) conjugated anti-mouse (Molecular Probes). Fluorescence images were collected with an Olympus IX-71 inverted fluorescence microscope with a Photometrix CoolSnapHQ CCD camera driven by DeltaVision software (Applied Precision, Seattle, WA). Collected images were deconvolved using Softworx deconvolution software (Applied Precision, Seattle, WA). For all images, 15 cycles of enhanced ratio deconvolution were used.

Electron microscopy

For conventional transmission electron microscopy cells were washed in buffer A or Ringer buffer, pH 7.2, fixed in 2.5% glutaraldehyde, 4% paraformaldehyde in 0.1 M cacodylic acid buffer, post-fixed in 1% OsO4 plus 0.8% ferrocyanide and 5 mM CaCl2 in 0.1 M cacodylic acid buffer for 30 min, dehydrated in acetone series and embedded in Polibed 812 epoxide resin. Sections of 70 nm were obtained and stained for 40 min in 5% aqueous uranyl acetate and for 5 min in lead citrate pH 12.0. Observation was made in a Zeiss 900 transmission electron microscope operating at 80 kV.

Electron tomography

Four hundred nanometre sections of epoxide embedded tachyzoites were obtained, collected in 200 mesh copper grids and stained as above. Tomographic tilt series over a range of 120° (±60°) by 1° angular increments were collected in a Tecnai 20 transmission electron microscope (FEI company, Eindhoven, the Netherlands) operating at 200 kV. Images were recorded on a bottom mounted Eagle 4 K CCD camera. Alignment of the tilt series was performed by cross correlation using the Inspect 3-D software package (FEI company). 3-D reconstruction was calculated by weighted back projection. Segmentation and generation of a 3-D model was calculated using IMOD software (Kremer et al., 1996).

Immunoelectron microscopy

For single localization studies by cryoimmunoEM, infected cells were fixed in 4% paraformaldehyde/0.05% glutaraldehyde (Polysciences, Warrington, PA) in 100 mM PIPES buffer. Samples were then embedded in 10% gelatin and infiltrated overnight with 2.3 M sucrose/20% polyvinyl pyrrolidone in PIPES at 4°C. Samples were frozen in liquid nitrogen and sectioned with a cryo-ultramicrotome. Sections were probed with purified rabbit anti-TgVP1 antibody (1:100) followed by secondary antibody conjugated to 18 nm colloidal gold, stained with uranyl acetate/methylcellulose and analysed by transmission EM.

For double localization studies, extracellular T. gondii were washed twice with PBS before fixation in 4% paraformaldehyde (Electron Microscopy Sciences, PA) in 0.25 M HEPES (pH 7.4) for 1 h at room temperature, then in 8% paraformaldehyde in the same buffer overnight at 4°C. Parasites were pelleted in 10% fish skin gelatin and the gelatin-embedded pellets were infiltrated overnight with 2.3 M sucrose at 4°C and frozen in liquid nitrogen. Ultrathin cryosections were incubated in PBS and 1% fish skin gelatin containing anti-TgCPL or TgVP1 antibodies at 1:600 and 1:200 dilutions, respectively, and then exposed to the secondary antibodies that were revealed with protein A-gold conjugates.

Acknowledgements

We thank Melina Galizzi and Cuiying Jiang for excellent technical assistance. We would like to specially thank Dr Vern Carruthers (University of Michigan) and Dr Fabiola Parussini for sharing unpublished information and reagents, for extensive discussions and critically reading the manuscript. We also would like to thank William Sullivan (University of Indiana) and Boris Striepen (University of Georgia) for Toxoplasma expression plasmids. Technical assistance with cryoimmunoEM (Fig. 1C) was provided by Wandy Beatty, Microbiology Imaging Facility (Washington University). We also thank Lia Carolina Medeiros for helping with movie renderization from the tomography data. This work was supported by U.S. National Institutes of Health Grant AI-079625 to S.N.J.M, and AI-034036 to L.D.S. K.M. was supported by a training grant from the Ellison Medical Foundation to the Center for Tropical and Emerging Global Diseases, Fundação Carlos Chagas Filho de Amparo a Pesquisa do Estado do Rio de Janeiro (FAPERJ) and Programa Jovens Pesquisadores CNPq-Brazil. D.A.P. was partially supported by an NIH T32 training grant AI-60546 to the Center for Tropical and Emerging Global Diseases. R.C. was supported in part by NIH research supplement 3R01AI068647-04S1. The name ‘plant-like vacuole’ or PLV was a recommendation of the participants of the 10th International Congress on Toxoplasmosis, in Kerkrade, the Netherlands, 19–23 June 2009.

Ancillary